Skip to main content

Biomarker candidate discovery

Let candidates emerge from papers, cross-reference 50+ databases, score GRADE tiers, and pool meta-analyses. Wet-lab experiments remain your responsibility.

The pipeline

Evidence in Motif, experiments in the lab

Search, extract, cross-reference, and score in one conversation. For the full discovery-to-validation path after literature triage, read our blog on biomarker discovery and validation to learn more.

Plain-language input

Ask a disease-focused question,
candidates emerge from papers

You do not need to pre-specify biomarker names. Motif searches PubMed, PMC, and Europe PMC, screens on title and abstract, and extracts diagnostic, prognostic, and predictive associations from full text. When hypotheses point toward druggable targets, read our blog on target identification to learn more about the path from literature to therapeutic programs.

  • MeSH-aware boolean queries rendered from your research objective
  • Full-text retrieval via PMC, Europe PMC, Unpaywall, and direct PDF upload
  • Expand search to add papers without restarting the conversation

Hi Alex, what would you like to explore?

Diagnostic biomarkers for early melanoma
Cross-reference BRAF V600E across databases
Prognostic proteins in idiopathic pulmonary fibrosis
Meta-analysis on CA-125 in ovarian cancer
Association Library
24found
Search...
Filters
Export

1. BRAF V600E mutation is diagnostic for melanoma vs benign nevi (sensitivity 92%, specificity 88%, case-control n=412)

↓ Supporting Evidence (3)

High
↓ View⎘ Copy

2. S100B elevation is prognostic for melanoma recurrence in discovery cohort (HR=2.1, n=198) and validation cohort (HR=1.9, n=143)

↓ Supporting Evidence (2)

Moderate
↓ View⎘ Copy

3. MMP7 overexpression is associated with IPF progression in prospective cohort (HR=1.74, 95% CI: 1.21–2.50, n=256)

↓ Supporting Evidence (2)

Moderate
↓ View⎘ Copy

Association library

Extract biomarker candidates
from full text

Motif emits structured association sentences with effect sizes, study design, and supporting quotes. Discovery and validation cohorts are tracked as separate associations when papers report independent replication, labeled as "discovery cohort" and "validation cohort" with their own sample sizes.

  • Each association gets a Motif Evidence Certainty tier (GRADE-adapted)
  • Diagnostic, prognostic, predictive, and mechanistic predicates supported
  • Org-wide Association Library with search, filters, and export

Cross-referencing

Map biomedical entities to 50+
external databases

After extraction, Motif cross-references biomedical entities against clinical, genomic, and regulatory sources including UniProt, HGNC, ClinVar, gnomAD, CIViC, ChEMBL, Open Targets, FDA, and more. Results appear in the Cross-Reference tab as per-database field panels, not as boolean chips on each association row.

  • Canonical IDs resolved across 50+ integration modules
  • Clinical sources include CIViC, DGIdb, ClinVar, and PharmGKB
  • Regulatory context from FDA drug and biomarker records where available

Cross-reference

BRAF V600E (gene)

Clinical · 3 databases matched

CIViCExpand

Evidence level: B

Clinical significance: Predictive

Disease: Melanoma

ClinVarExpand

Review status: Expert panel

Classification: Pathogenic

Variant: BRAF p.V600E

FDAExpand

Drug: vemurafenib

Indication: BRAF V600E mutant melanoma

Approval: 2011

PD-L1 → pembrolizumab PFS

Hazard ratio · 4 studies · N=1,842

GRADE: moderate
Hugo 2016
0.580.41–0.82
Rizvi 2017
0.710.52–0.97
Hellmann 2018
0.630.49–0.81
Pooled
0.640.54–0.76
38%τ² 0.044 studies assessed

Pooled evidence

Meta-analysis when ≥3
comparable studies exist

When enough studies report comparable effect sizes, Motif pools estimates with forest and funnel plots, RoB-2 coloring per study, and a pooled GRADE certainty rating distinct from per-association scoring. Download the full workbook (.xlsx) with the GRADE sheet. Read our blog on how AI is revolutionizing biomarker discovery to learn more.

  • Per-association GRADE-adapted tiers: High, Moderate, Low, Very Low
  • Help icon opens the Motif Evidence Certainty methodology dialog
  • Meta-analysis requires at least three studies with comparable effect sizes

Grant-ready outputs

Export evidence for
your next experiments

Export per pipeline: articles, associations, cross-references, meta-analyses, your report, and search provenance, or download a Word manuscript with APA 7 or Vancouver citations. Use the export to brief wet-lab teams; Motif does not run assays or file regulatory submissions.

  • Excel or CSV with sheets per result type
  • Word ZIP with reference list and BibTeX
  • JSON for downstream analysis in R or Python

Export Results

Choose a format to export your research results

Export Format
JSON
{ }
JSON
Structured data for API integration
.json
Reference Style
Citation Style
Results to include
Fluid biomarkers for PD diagnosis
Pipeline

Biomarker candidate discovery FAQ

What Motif covers in candidate discovery vs what stays in the lab

What does Motif do in biomarker discovery?

Motif searches PubMed, PMC, and Europe PMC, extracts structured biomarker associations from full text, cross-references biomedical entities against 50+ databases, scores each association with a GRADE-adapted certainty tier, and can pool meta-analyses when at least three comparable studies exist. Biomarker candidates emerge from the literature rather than a pre-loaded list.

Does Motif run analytical or clinical validation?

No. Analytical validation (assay precision, sensitivity, specificity) and clinical validation (patient outcomes, utility) happen in your lab and trials. Motif surfaces published evidence, including discovery vs validation cohort reporting, to inform which candidates merit wet-lab follow-up.

How does Motif score evidence quality?

Each association gets a Motif Evidence Certainty score based on study design, sample size, and reporting quality, surfaced as High (≥0.85), Moderate (≥0.60), Low (≥0.35), or Very Low. The help dialog explains the GRADE-adapted methodology. Pooled meta-analyses receive a separate GRADE rating with downgrade reasons in the meta-analysis card.

How are discovery and validation cohorts handled?

When a paper reports independent discovery and validation cohorts, Motif extracts them as separate associations: one for the discovery cohort and one for the validation cohort, each with its own sample size and effect size. This lets you compare replication evidence before committing assay development resources.

When is meta-analysis available?

Meta-analysis runs when at least three studies report comparable effect sizes for the same biomedical entity and outcome context. Motif produces a forest plot, funnel plot, I² and τ² statistics, per-study risk-of-bias coloring, and a pooled GRADE certainty rating. Export the full workbook (.xlsx) for your records.

What are the limitations?

Motif screens on title and abstract before reading full text. Paywalled PDFs without open access may not be retrievable. Cross-reference coverage depends on database availability for each biomedical entity. Scope depends on your plan: Starter searches up to 5 papers per query, Pro up to 40. Researchers should verify key findings in the source papers before validation investments.

Start navigating biomarker evidence

Join researchers using Motif to find and score biomarker evidence from published literature.

Limited early access for researchers. No spam.

Biomarker candidate discovery - Motif